A great deal of popular attention has been given to the Mozart effect—an increase in spatial ability following listening to Mozart. Three hypotheses have been advanced to explain this association: Mozart priming the neural pathways used for spatial reasoning, Mozart generally increasing mood and arousal and thus performance, or individuals' preference for Mozart, a different form of music, or even silence leading to an optimal mood for test-taking. The current study sought to differentiate among these three hypotheses. Data were collected from 41 college students (20 male, 21 female) assessed on a spatial relations subtest from the Stanford-Binet following exposure to either music or silence. Participants self-reported how awake they felt and their preference for their particular condition. Results indicated a positive effect of listening to Mozart, although arousal mediated this association. No effect of preference was evident. Implications for theory and application are discussed.